ON THE KINETICS OF HYDRATE FORMATION
IN THE SYSTEM CO2-H2O
A. Ullrich and R. Eggers
Technical University Hamburg-Harburg, Thermal Separation Processes
Eissendorfer Str. 38, 21073 Hamburg, Germany
e-mail: [email protected], FAX: ++49-(0)40-42878-2859
The formation of hydrate during the isochore cooling and along the pressure release was
studied. In a previous work it was found that the pressure release of wet CO 2 is significantly
slower than the release of dry CO2 . The reason is the formation of hydrate in the venting
line. The objective of this work is the evaluation of the blocking risk of pipelines by hydrate
crystals in processing steps. Experiments were carried out in a 87 ml view-cell to observe
visible hydrate formation since crystals of that size can cause the mentioned problems. In
a second part the influence of a contact surface on the hydrate formation was studied. The
investigated materials were steel, glass and PTFE. The starting conditions of all experiments
were within a temperature range of T = 280 - 285 K and a pressure range of p = 6 - 20 MPa.
As a result of the experiments, the hydrate formation of visible crystals can be refered to two
major criteria (a) the water content of the CO2 and (b) the degree of subcooling. Visible hydrate
formation can be observed at a minimum subcooling of T = 5 K. If the water content at the
beginning of the experiment is below 0,05 wt-% no hydrate formation can be observed. The
surface properties of the material have no direct influence on the hydrate formation process.
But during the pressure release hydrate crystals detach from the PTFE surface. In contrast
to that the hydrate crystals do not detach from the steel surface and the glass surface. Thus,
the adhesion between hydrate crystals and steel and hydrate crystals and glass, respectively, is
stronger than between hydrate crystals and PTFE.
INTRODUCTION
In many industrial processes where carbon dioxide is used as a solvent small amounts of water
dissolve in the carbon dioxide. The water is mostly introduced with the feed material. During
process-related cooling or pressure release and also during an emergency discharge the phase
behaviour of the system CO2 -H2 O changes. Under certain conditions this leads to CO2 -hydrate
formation. Gas hydrates belong to the group of clathrates. The water molecules build lattice
structure with specific cavities. These cavities are occupied by gas molecules. The included
gas molecules stabilise the structure to form an ice-like compound [1]. The problem of hydrate
formation can also appear during the piping of liquid CO 2 , that contains small amounts of water.
Especially in the suction line of pumps there is a risk of hydrate formation due to the pressure
drop or changes in the temperature. The arising solid phase can interfere the flow. It can
cause blockages of pipes and mechanical damages to certain plant components like membrane
pumps. In a previous work it was found that the pressure release of wet CO 2 is significantly
slower than the release of dry CO2 . The reason is the formation of hydrate in the venting line
[2]. The objective of this work is to investigate the kinetics of hydrate formation during cooling
and during pressure release of wet CO2 .
EXPERIMENTAL
For the experiments a 87 ml view cell is used. Figure 1 shows the experimental set-up. The
maximum operating pressure of the cell is pmax = 50 MPa and the maximum operating temperature is T = 400 °C.
Figure 1: Experimental set-up
CO2 is conveyed through a water filled vessel and than filled in the view cell. Different thermodynamic starting conditions were adjusted. To ensure equilibrium between CO 2 and H2 O the
experiments started about five hours after adjusting the starting conditions for the experiments.
In a first set of experiments the view cell is cooled down to investigate an isochore system. In
a second set the hydrate formation during the pressure release is observed. All starting conditions are in the area of the phase transition between LCO2 -LH2 O region and H-LH2 O region (see
figure 2). Finally in a last set of experiments the influence of different contact materials is to
be analysed. Therefor a plate of steel, glass and PTFE are put into the view cell, respectively.
Then the system is cooled down and the processes in the cell are observed. During all the
experiments temperature and pressure are recorded. The temperature is measured with a thermocouple that is placed in the centre of the view cell. The pressure is measured via a pressure
gauge. The processes in the view cell are recorded by a CCD-video system. To identify the
conditions the pictures are recorded. A video timer is installed that is synchronised with the
data recording of pressure and temperature.
9
8
7
GCO
LCO
LH O
H
I
2
Pressure [MPa]
2
6
2
: gaseous CO2-rich phase
: liquid CO2-rich phase
: liquid H 2O-rich Phase
: CO2-Hydrate
: Ice (H 2O-rich)
LCO - LH O
2
2
5
H - LCO
4
3
2
1
0
220
v ap
our
p re s
s u re
ep
c u rv
ure
2
G CO - LH O
CO 2
2
H - GCO
2
283 K
2
I-G
240
260
280
300
Temperature [K]
Figure 2: -phase diagramm of the system CO2 /H2 O (according to [3])
30
28
26
24
22
Pressure [MPa]
20
18
GCO
LCO
LH O
H
I
2
2
2
: gaseous CO2-rich phase
: liquid CO2-rich phase
: liquid H 2O-rich Phase
: CO2-Hydrate
: Ice (H 2O-rich)
1
LCO - LH O
2
2
16
14
2
12
H - LCO2
10
3
8
6
4
H - GCO
2
0
220
230
240
250
260
270
Temperature [K]
GCO - LH O
2
2
2
280
I - GCO
290
300
2
Figure 3: Typical development of pressure over temperature for a cooling experiment in the
-phase diagramm for the system CO2 /H2 O (according to [3]). Points 1, 2 and 3 mark the
conditions when the pictures shown in figure 4 were taken.
RESULTS
Figure 3 shows a typical development of pressure over temperature for a cooling experiment.
The dots in the diagram mark the conditions when the pictures shown in figure 4 were taken.
Hydrate formation could be observed when the starting pressure was above p = 6 MPa and the
temperature was T = 285 K. Visible hydrate formation could be observed when the system is
subcooled by T = 6 K. Further cooling leads to an increase of the crystal size. When reaching
a certain size the formed crystals sink down to the surfaces. On the surface they eventually
build a hydrate layer with an ice-like structure.
In the pressure release experiments the temperature in the view cell drops independently from
the initial conditions beneath 283 K when the valve is opened. It takes about 10 seconds to reach
the H-LCO2 -GCO2 phase-boundary. Than it takes another 5 minutes to reach ambient pressure.
Figure 5 shows a typical development of pressure over temperature during a pressure release
experiment. The points mark the conditions when the picture shown in figure 6 were taken.
a)
t=0s
b)
t = 55 min 64 s
c)
t = 90 min 55 s
Figure 4: Surface (in this case: PTFE) prior and after hydrate formation. T = T = 285 K,
p = p = 20,0 MPa, w = 0,14 wt-%; T = 275 K, p = 11,5 MPa, w = 0,095 wt-%;
T
= 273 K, p
= 9,8 MPa, w = 0,085 wt-% (The water content is estimated according to
figure 7)
As in the cooling experiments visible hydrate formation can be observed when the system
reaches a subcooling of T = 6 K. The formed hydrate crystals have fluffy snow-like structures.
Experiments with varying contact materials show no differences in the hydrate formation itself.
In all cases some small hydrate particles form initially. They sink to the surface and finally,
grow to a hydrate layer. However, when the system is depressurised a difference occurs. In the
experiments using a PTFE plate the hydrate crystals detach from the surface and are carried
away with the flow (see figure 6). In the experiments with the glass plate and the steel plate
none of that can be seen.
10
9
8
GCO
LCO
LH O
H
I
2
2
2
Pressure [MPa]
7
6
: gaseous CO2-rich phase
: liquid CO2-rich phase
: liquid H 2O-rich Phase
: CO2-Hydrate
: Ice (H 2O-rich)
5
LCO - LH O
2
2
H - LCO2
4
1
3
2
2
H - GCO
1
0
220
GCO - LH O
2
230
240
250
2
260
270
Temperature [K]
280
I - GCO
290
2
300
2
Figure 5: Development of pressure over temperature for a fast pressure release experiment.
The points mark the conditions when the picture shown in figure 6 were taken.
a)
b)
Figure 6: Behaviour of hydrate crystals on a PTFE plate during the pressure release of wet
CO2 . (T =275 K, p =3,4 MPa; T =269 K p =2,9 MPa)
DISCUSSION
During the pressure release a high supersaturation is achieved very fast. This causes the formation of many small hydrate particles. These small particles agglomerate to a more fluffy
snow-like structure. The small hydrate particles are partially carried away with the flow and can
accumulate at deflection units or bottlenecks causing strictures or even blocking of pipelines.
At fast decreasing of pressure and coupled decreasing of temperature the solubility of water
in CO2 decreases. Thus, strong nucleation, but delayed by supersaturation, occurs since the
conditions are in the hydrate region of the phase diagramm. Due to the fast change of solubility and the coupled high supersaturation the nucleation rate significantly dominates against the
growth of hydrate crystals.
The cooling of the systems leads to decreasing solubility of water in CO 2 . For example, in
the experiment with a starting pressure of p = 20 MPa and a starting temperature of T = 285 K
the water content is w = 0.14 wt-%. After 3300 seconds the first hydrate crystals are visible.
The pressure at this time is p = 11.5 MPa and the temperature is T = 275 K. The water content
at this point is w = 0,095 wt-%. Thus, the cooling of the system first leads to a supersaturation
that results in the precipitation of hydrate crystals. Since hydrate if of higher density than the
surrounding fluid, the crystals sink to the surface were their size further increases. In literature three types of macroscopic structures of hydrate crystals formed from an aqueous solution
are described. Aya et al. [5] differentiate the frost-type, the tree-type and the wool-type. The
hydrate formed on the surfaces in this experiments from a CO2 -rich solution with dissolved
water corresponds to the frost-type. In contrast to the fast pressure release with the fast change
in temperature, here the growth of crystals dominates the nucleation rate. Thus, in this case
more dense structures can develop. In experiments with an starting pressure of p = 6 MPa and
a starting temperature of T = 285 K no hydrate formation could be observed. In this case the
water content at the beginning of the experiment is w = 0,05 wt-%. By cooling the system to
T = 268 K (p = 3.1 MPa) the water content decreases by w = 0.3 10 3 . The change in the water content is not sufficient to induce visible hydrate formation.
It was found in these experiments that the hydrate crystals detach from PTFE surfaces while
they stick to glass and steel surfaces. This happens when a sufficient turbulence is present.
Thus, the hydrate crystals are stronger bound to the glass and the steel surface than to the PTFE
surface. In contrast to PTFE, glass and steel show a hydrophilic surface. Therefore the molecular attractive forces are significant stronger in the case of steel/hydrate and glass/hydrate.
Raraty and Tabor [4] measured the adhesion strength of ice to different materials. For the
temperature range between -10 °C and -30 °C they
found a nearly constant value for the ad
hesion strength of ice to steel of about S = 20 . In the case of ice and PTFE they found a
0.32
0.30
T
T
T
T
0.28
0.26
0.24
0.22
=
=
=
=
244
263
280
298
K
K
K
K
T
T
T
T
=
=
=
=
255
272
291
304
K
K
K
K
w [wt-%]
0.20
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
0
2
4
6
8
10
12
14
16
18
20
22
Pressure [MPa]
Figure 7: Solubility of water in CO2 (according to [6])
adhesion strength of S = 1.5 . Considering the substantial similarity of ice to hydrate it is
possible to transfer these values to the hydrate systems. The adhesion strength between steel
and hydrate and PTFE and hydrate differ by one order of magnitude. Since steel and glass have
similar properties concerning their hydrophilic character it can be assumed that the values of
the adhesion strength between glass and hydrate are close to the values for steel and hydrate.
CONCLUSIONS
As a result of the experiments the hydrate formation of visible crystals can be refered to two
major criteria (a) the water content of the CO2 and (b) the degree of subcooling. All experiments show that for observing visible hydrate formation in a reasonable time a certain degree
of subcooling is essential. In the experiments the values of the subcooling were in an range of
T = 5 - 8 K. Fast changes of solubility result in a more fluffy snow-like structure while slower
changes lead to a more dense ice-like structure. In the first case the nucleation rate dominates while in the second case the growth of crystals is more dominating. Overall the hydrat
formation occurs in the bulk phase as homogeneus nucleation. No preferred spots of hydrate
formation could be found.
The experiments with different water contents at the beginning of the experiments indicate that
with a water content below w = 0.05 wt-% no visible hydrate formation can be observed. Thus,
with that water content the risk of the blocking of venting lines is minimal.
The surface properties of the material have no direct influence on the hydrate formation process. But during the pressure release hydrate crystals detach from the PTFE surface. In contrast
to this the hydrate crystals do not detach from the steel surface and the glass surface. Thus the
adhesion between hydrate crystals and steel and hydrate crystals and glass is stronger than between hydrate crystals and PTFE. Considering measurements of the adhesion strenght of ice
to different materials [4] the differences of the adhesion strength between hydrate/steel and
hydrate/teflon can be estimated to one order of magnitude.
ACKNOWLEDGEMENT
We thank the Max-Buchner Research Foundation for their support (key number 2206).
REFERENCES
[1] SLOAN, E. D., Clathrate Hydrates of Natural Gases, Marcel Dekker Inc., New York, 1990
[2] FREDENHAGEN, A., EGGERS, R., Einfluss der Hydratbildung auf die Druckentlastung
von feuchtem CO2, Chemie Ingenieur Technik, 72(10), 2000, pp. 1221 - 1224
[3] GMELIN, L., Gmelin Handbuch, 8 Edition, Kohlenstoff Part C3, Verlag Chemie, Weinheim, 1973, pp. 36 ff.
[4] RARATY, L. D., TABOR, D., The adhesion and Strength Properties of Ice, Proceedings
of the Royal Society of London, A245, 1958, pp. 184 - 201
[5] AYA, I., YAMANE, K., NARIAI, H., Solubility of CO 2 Hydrate at 30 MPa. Energy,
Vol. 22, No. 2/3, 1997, pp. 263 - 271
[6] SONG, K. Y., KOBAYASHI, R., Water content of CO 2 in equilibrium with liquid water
and/or hydrates. Society of Petroleum Engineers, Formation Evaluation, 1987, pp. 500 508